In fiber optic technology light signals are used in place of electrical signals to transmit information from one place to another. In the place of metal wires, transparent optical fibers carry light signals bearing the transmitted information.
One of the desired components in fiber optic technology is a low cost, high-performance coupler by which light signals sent on one optical fiber may be divided or split into light signals on several optical fibers. Conversely light signals on several optical fibers may be combined into one fiber by such a coupler. In the more common electrical technology this is a straightforward matter. Wires may be spliced together to form multiple branches. Electrical signals sent along one wire naturally propagate along all branches connected to the wire; the converse situation is also true.
In fiber optic technology the splitting and combining of the light signals is a much more complicated matter. Connections must be made very carefully to ensure that the intensity of the light signals are shared equally among the fiber branches and that the loss in intensity is minimized. Furthermore, any resulting phase shifts in the light signals must be accounted for. Hence various schemes for optical splitters have been devised.
Heretofore, integrated optical waveguides on LiNbO.sub.3 and glass substates, fused fiber optic star couplers, slab wave guides, active couplers (either acting as combiners or splitters), optical hologram or Fourier transform fiber optic splitters, and GRIN lens fiber optic splitters have been investigated. However, all of these approaches have had deficiencies of performance, such as in excess loss, loss deviation, frequency response, phase preservation and expandability of the splitting ratio, and costs, such as in fabrication of the splitter or interconnection with the rest of a fiber optic network.
For example, previous fiber optic splitters, such as fused splitters, have had splitting ratios of 1:N where N is invariably an even number. Where an odd number of second fibers is desired, the fused splitter wastes the precious energy in the undesired output optic fiber. Another example is that previous GRIN lens fiber optic couplers typically have had a 1:2 splitting ratio and have had difficulty with ratios greater than 1:16.
Thus what is needed is a reliable and low cost fiber optic coupler having a splitting ratio of 1:N where N can be an arbitrary number and where N can be a large number, i.e., greater than 16. The present invention provides for such a fiber optic coupler.